INTERPRETING COPD DYSPNEA CHANGE: SENSITIVITY, RESPONSIVENESS, AND PREDICTIVE VALIDITY OF THE DMQ-CAT

Dyspnea, or uncomfortable labored breathing, is an important patient-reported outcome (PRO). It is the primary and most disabling symptom of COPD. There is a high priority to improve dyspnea PRO assessment and predict exacerbations. Frequent exacerbations are associated with increased disability, decreased quality of life (QOL), and accelerated lung function decline. Goals: To test the relative sensitivity to change, responsiveness and predictive validity of a comprehensive dyspnea outcome computer adaptive test (CAT) that measures new anxiety and activity avoidance domains and is more efficient to administer than existing dyspnea scales. Expected Outcomes: We expect to: (1) begin to transform how dyspnea is assessed; (2) improve dyspnea symptom management; (3) impact functional status; (4) improve QOL; (5) facilitate the earlier treatment and prevention of exacerbations; (6) improve COPD prognosis and survival; and (7) improve COPD healthcare utilization. Population to be Served: Adults with Chronic Obstructive Pulmonary Disease (COPD). Theoretical Background: COPD research and practice has focused predominately on evaluating and treating the sensory component of dyspnea. However, dyspnea (like pain) is a complex symptom with sensory and affective components. Current dyspnea scales are limited in scope by only measuring the sensory component of dyspnea, and have high respondent and data collection burden and/or high measurement error. Only sensory dyspnea to date has been used to classify COPD severity. The DMQ-CAT addresses the limitations of other dyspnea scales by using state-of-the-art item banking method techniques to provide a tailored multidimensional dyspnea assessment using only a minimal number of items. The 71-item DMQ-CAT captures four distinct dyspnea constructs: intensity, anxiety, activity avoidance, and self-efficacy to evaluate outcomes of COPD pharmacologic, pulmonary rehabilitation, and cognitive-behavioral therapy. The DMQ-CAT showed evidence of minimal ceiling and floor effects, acceptable Rasch model fit statistics, high internal consistency reliability, good accuracy compared to the total item pool, and good to excellent concurrent and known-groups validity. The relative sensitivity to change, responsiveness, and predictive validity of the DMQ-CAT is not currently known. Hypotheses: The DMQ-CAT has superior sensitivity to change and responsiveness compared to other widely used COPD dyspnea scales. The DMQ-CAT has superior predictive validity compared to other dyspnea measures to predict acute COPD exacerbation incidence and severity. Study Design: To achieve Aims 1 and 2, we will use a prospective observational design and a target sample of 110 pulmonary rehabilitation (PR) outpatients with COPD. We will use distribution-based statistics to compare the sensitivity to change of dyspnea instruments including: effect sizes, standardized response means (SRM), SEM (standardized error of measurement), and the percentage of participants exceeding the minimal detectable change (MDC90) from baseline to discharge from PR. We will estimate the minimal clinically important difference (MCID) for the DMQ-CAT and compare the percentage of participants exceeding MCID thresholds using established MCIDs for dyspnea instruments. For Aim 3, we will use a retrospective observational design to conduct secondary analyses from existing data of 240 patients with COPD. We will use odds ratios (OR) and 95% CI and the area under the receiver operator curve (ROC) to compare the predictive validity of different dyspnea measures.